Monocular Occupancy Prediction for Scalable Indoor Scenes
This addresses the problem of scalable indoor scene analysis for robotics or AR/VR applications, but it is incremental as it adapts outdoor methods to indoor settings.
The paper tackles monocular 3D occupancy prediction for indoor scenes, which is less explored than outdoor driving scenes, by proposing the ISO method with a novel D-FLoSP module and introducing the large-scale Occ-ScanNet benchmark. It achieves state-of-the-art performance on NYUv2 and Occ-ScanNet, with Occ-ScanNet being 40 times larger than NYUv2.
Camera-based 3D occupancy prediction has recently garnered increasing attention in outdoor driving scenes. However, research in indoor scenes remains relatively unexplored. The core differences in indoor scenes lie in the complexity of scene scale and the variance in object size. In this paper, we propose a novel method, named ISO, for predicting indoor scene occupancy using monocular images. ISO harnesses the advantages of a pretrained depth model to achieve accurate depth predictions. Furthermore, we introduce the Dual Feature Line of Sight Projection (D-FLoSP) module within ISO, which enhances the learning of 3D voxel features. To foster further research in this domain, we introduce Occ-ScanNet, a large-scale occupancy benchmark for indoor scenes. With a dataset size 40 times larger than the NYUv2 dataset, it facilitates future scalable research in indoor scene analysis. Experimental results on both NYUv2 and Occ-ScanNet demonstrate that our method achieves state-of-the-art performance. The dataset and code are made publicly at https://github.com/hongxiaoy/ISO.git.